UcD400 Q & A

[pburke]

Originally posted by classd4sure



So far Nichicon are the cheapest I've found that might be good quality, screw in terminals, and are dirt cheap

Do quality caps in the psu still benefit from bypassing with a small value high quality cap? I don't want to have to make a PCB..

Thanks

Nichicon makes a lot of different grades. The "good stuff" for audio use in that voltage rating isn't "dirt cheap" at all...

check page 8 of this catalog:

http://www.percyaudio.com/Catalog.pdf

for the KG gold or KG Super Series caps. 10,000uF 80V go for $36.95

still ages better than Blackgates, and yes, bypassing is very much a possibility, just as paralleling many smaller values is. I'd use a 10k, a 1k, a 22uF and maybe even a 0.1 Auricap to top it off, all depending on the budget. My current SS amps are filled with Wimas on every 10,000 uF cap inside (and that's not a DIY product)

Peter

[kuribo]

Curious as to why you would add all the bypass caps after Bruno said they were unnecessary????

[classd4sure]

Hi,

You're right Peter they aren't exactly dirt cheap, cheaper than most but all the same, that's the same place I was looking at originally as well, the Gold series are almost half the price though. I do have a place where caps are dirt cheap, example SMH series by UCC 12,000uF 100Vdc $11.82 Canadian, how's that for a steal? I'm going to keep looking though.

Kuribo, I wouldn't If it's redundant it's redundant. It's good to have people sharing some info though. I would by pass the bridge rectifiers however, 1 cap per diode, and add some bleeder resistors.

Thanks

[Bruno Putzeys]

Originally posted by classd4sure
Kuribo, I wouldn't If it's redundant it's redundant. It's good to have people sharing some info though. I would by pass the bridge rectifiers however, 1 cap per diode, and add some bleeder resistors.

Bypassing the bridge rectifier is a MUST. On one of my first designs, emission was wholly and completely dominated by the 50Hz rectifier!
(More precisely that was during the prescan which is done using peak detection. When qpk detection was used to recheck the overs it was OK, but that's entirely owed to the leniency of quasi-peak detection).

[classd4sure]

Hi Jan-Peter,

It's been said a few times the 63 volt limit of the module was imposed by those 63V 470uF caps, since they're now being changed to 100V caps, will the overvoltage protection stay ~63V?

If it will stay at 63V would you recommend anyone try to change or ajust it? If so what would be required to make such a change, and what level should it never exceed in order to still keep it useful in protecting the module? Or would you possibly consider making that kind of help available through private email should the situation arise?

I ask because it might be nice if someone like myself were to try and hit close to 63Vdc worst possible case and it wound up being just over that, a quick adjustment or two would be cheaper than buying another transformer. We'd of course, neeever, attempt to make it a UcD500.

Plitron state a 4% load regulation for all their transformers from 500VA to 1500VA, I have to question how close they achieve that mark in reality. Anyway, it would take the pressure off of selecting the transformer to get the most out of the module.

Thanks

[Jan-Peter]

Hi Chris,

The overvoltage wil not be changed and sta at 62-63V .

Everbody can ofcourse mail me private for "how to disable the overvoltage protection"

Regards,

Jan-Peter

www.hypex.nl

[classd4sure]

Hi,

Thanks for the response.

Disable huh? 62~63..Hmmmm 35Vac it shall be!

Regards

[analogspiceman]

Originally posted by Bruno Putzeys
Bypassing the bridge rectifier is a MUST. On one of my first designs, emission was wholly and completely dominated by the 50Hz rectifier! (More precisely that was during the prescan which is done using peak detection. When qpk detection was used to recheck the overs it was OK, but that's entirely owed to the leniency of quasi-peak detection).

This is very true. Rectifier noise arises because, with no power factor correction, the ac rectifiers only conduct at the peaks of the ac input voltage. When they turn off the voltage on their line side takes a sudden, sharp step. This generates a long (to many MHz) picket fence of conducted EMI noise spikes spaced at 100Hz. Adding a pair of 10nF to 100nF capacitors across the bridge rectifier (from ac terminals to minus) stops this step and knocks down the picket fence. I've found that using caps across all four rectifiers rather than only across the two bottom ones actually worsens overall EMI, probably because the plus terminal of the rectifier bridge is generally noisier with switching frequency differential mode EMI than is the minus terminal.

Regards -- analog(spiceman)

[jam]

Spiceman,

Is this also true when using Hexfreds?

Regards,
Jam

[classd4sure]

Hi

AnalogSpiceman, may I be so bold as to request you post a quick diagram of your recommended bypass method? I'd appreciate it.

I guess soft recovery would be a beneficial feature.

There's already alot of threads on this from what I see and it's kind of off topic, though still towards the cause, but, they seem to agree the rectifiers have a some sort of sonic effect, hard to believe, is this more voodoo? Since I highly value the opinion of people here I thought I'd ask for some opinions on this.

I'm currently looking at the basics I didn't think I'd have to start looking for the golden rectifier that sounds the best as well.

Thanks